CN118376948B - Performance measurement method of load point power supply - Google Patents
Performance measurement method of load point power supply Download PDFInfo
- Publication number
- CN118376948B CN118376948B CN202410815436.XA CN202410815436A CN118376948B CN 118376948 B CN118376948 B CN 118376948B CN 202410815436 A CN202410815436 A CN 202410815436A CN 118376948 B CN118376948 B CN 118376948B
- Authority
- CN
- China
- Prior art keywords
- sequence
- time
- electrical signal
- electric signal
- taking
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000691 measurement method Methods 0.000 title abstract description 6
- 238000000034 method Methods 0.000 claims abstract description 74
- 238000005259 measurement Methods 0.000 claims abstract description 45
- 238000001914 filtration Methods 0.000 claims abstract description 43
- 230000008569 process Effects 0.000 claims abstract description 35
- 238000010586 diagram Methods 0.000 claims abstract description 27
- 230000007774 longterm Effects 0.000 claims abstract description 26
- 230000005856 abnormality Effects 0.000 claims abstract description 15
- 230000002159 abnormal effect Effects 0.000 claims abstract description 10
- 238000011156 evaluation Methods 0.000 claims abstract description 7
- 230000008859 change Effects 0.000 claims description 31
- 108010076504 Protein Sorting Signals Proteins 0.000 claims description 9
- 230000001174 ascending effect Effects 0.000 claims description 9
- 238000013506 data mapping Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000009825 accumulation Methods 0.000 claims description 2
- 238000007689 inspection Methods 0.000 claims description 2
- 230000003044 adaptive effect Effects 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 8
- 238000005070 sampling Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000003064 k means clustering Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000013112 stability test Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/40—Testing power supplies
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing Electric Properties And Detecting Electric Faults (AREA)
Abstract
The application relates to the technical field of signal filtering, and provides a performance measurement method of a load point power supply, which comprises the following steps: collecting an electric signal; determining a ripple noise interference joint index of each time period based on an evaluation result of the interference degree of the extreme point in the time domain waveform diagram of the electrical signal time sub-sequence in each time period; determining a measured interference long-term disturbance coefficient based on a ripple noise interference joint index and a consistency characteristic of long-range autocorrelation among the electrical signal time sub-sequences; determining an electrical signal ripple abnormality index based on the ripple noise interference joint index and the measured interference long-term disturbance coefficient; determining a step factor of each iteration in the electric signal filtering process based on the electric signal ripple abnormal index of the electric signal time sequence; and obtaining a performance measurement result of the load point power supply based on the step factor by adopting an LMS filtering algorithm. The application optimizes the LMS filtering algorithm to filter the electric signal and improves the accuracy of the performance measurement result of the load point power supply.
Description
Technical Field
The application relates to the technical field of signal filtering, in particular to a performance measurement method of a load point power supply.
Background
The load point power supply is used as a secondary power supply of a power supply loop, and is generally used for supplying power to integrated digital circuits such as FPGA, CPU, CPLD, MCU due to higher power density and faster transient response speed of the load point power supply. However, as the degree of integration of integrated circuits increases, the requirements for the performance of the load point power supply increase, and it is often necessary to measure the power supply performance of the load point power supply in real time.
In order to measure the performance of the load point power supply, the stability of the current in the power supply loop is generally measured, and the stability of the power supply of the load point power supply is reflected by the stability of the current. In general, an integrated digital circuit has the characteristics of multiple pins, low voltage and high current, and in this case, the measurement of the current in the power supply loop of the load point power supply has larger interference of ripple noise, which affects the measurement of the performance of the load point power supply.
With the continuous development of digital signal processing technology, the interference of ripple noise on current signals in a power supply loop is often reduced by a digital filtering technology, so as to improve the accuracy of the performance measurement of a power supply at a load point. For example, the conventional adaptive filtering LMS (LEAST MEAN Square) algorithm is commonly used for the filtering process of current signals. However, the traditional LMS adaptive filtering algorithm selects a single fixed step factor, so that the compatibility of convergence speed and steady-state error in the current signal processing of a power supply loop is poor, the filtering performance of the algorithm is poor, and the performance of a load point power supply cannot be accurately measured.
Disclosure of Invention
The application provides a performance measurement method of a load point power supply, which aims to solve the problem that the measurement result is affected by poor electric signal filtering effect in the measurement process of the load point power supply by an LMS filtering algorithm, and adopts the following specific technical scheme:
The application provides a performance measurement method of a load point power supply, which comprises the following steps:
collecting an electric signal in a load point power supply loop;
Determining a ripple noise interference joint index of each time period based on an evaluation result of the interference degree of the extreme point in the time domain waveform diagram of the electrical signal time sub-sequence in each time period;
Determining a long-term disturbance coefficient of the measured interference in the load power supply measuring process based on the ripple noise interference joint index of all time periods and the consistency characteristic of the long-range autocorrelation among the electrical signal time subsequences;
determining an electrical signal ripple abnormality index of the electrical signal time sequence based on the ripple noise interference joint index of each time period and the measured interference long-term turbulence coefficient in the load point power supply measuring process;
Determining a step factor of each iteration in the electric signal filtering process based on the electric signal ripple abnormal index of the electric signal time sequence; and obtaining a performance measurement result of the load point power supply based on the step factor by adopting an LMS filtering algorithm.
Preferably, the method for determining the ripple noise interference joint index of each time period based on the evaluation result of the interference degree of the extreme point in the time domain waveform diagram of the time sub-sequence of the electric signal in each time period comprises the following steps:
Taking a sequence formed by all the collected electric signals according to the ascending order of time as an electric signal time sequence, and dividing the electric signal time sequence into a preset number of electric signal time subsequences in a time dimension in a uniform dividing mode;
Respectively taking each electrical signal time sub-sequence as input, and acquiring a time domain waveform diagram of each electrical signal time sub-sequence by utilizing MATLAB software;
determining a ripple noise interference crest index of each time period based on the stability degree of the euclidean distance between the crest points in the time domain waveform diagram of each electrical signal time sub-sequence;
determining a ripple noise interference trough index of each time period based on the stability degree of the euclidean distance between trough points in the time domain waveform diagram of each electrical signal time sub-sequence;
The ripple noise interference combined index of each time period consists of a ripple noise interference crest index and a ripple noise interference trough index, wherein the ripple noise interference combined index is respectively in positive correlation with the ripple noise interference crest index and the ripple noise interference trough index.
Preferably, the method for determining the ripple noise interference crest index of each time period based on the stability degree of the euclidean distance between the peak points in the time domain waveform diagram of each electrical signal time sub-sequence comprises the following steps:
Taking a set formed by a preset number of peak points with the minimum Euclidean distance with each peak point in a time domain oscillogram of each electric signal time sub-sequence as an adjacent peak point set of each peak point;
Taking the mean value of Euclidean distances between each peak point and all the peak points in the adjacent peak point set of each peak point as the variation characteristic value of each peak point, and taking a sequence formed by the variation characteristic values of all the peak points according to the sequence of time ascending as the peak variation characteristic value sequence of the time domain waveform chart of each electric signal time subsequence;
taking the data mapping result of the information entropy and standard deviation product of all elements in the wave crest variation characteristic value sequence of the time domain waveform chart of each electric signal time sub-sequence as a molecule;
taking the sum of the absolute values of the difference values of all adjacent elements in the wave crest variation characteristic value sequence of the time domain waveform diagram of each electric signal time subsequence and 1 as a denominator;
Taking the difference value of the ratio of 1 to the numerator and the denominator as the ripple noise interference crest index of the corresponding time period of each electrical signal time subsequence.
Preferably, the method for determining the ripple noise interference trough index of each time period based on the stability degree of the euclidean distance between the trough points in the time domain waveform diagram of each electrical signal time sub-sequence comprises the following steps:
Taking a set formed by a preset number of wave crest points with the minimum Euclidean distance with each wave trough point in a time domain oscillogram of each electric signal time subsequence as an adjacent wave trough point set of each wave trough point;
Taking the average value of Euclidean distances between each wave valley point and all wave valley points in a nearby wave valley point set of each wave valley point as a change characteristic value of each wave valley point, and taking a sequence formed by the change characteristic values of all wave valley points according to the sequence of time ascending as a wave valley change characteristic value sequence of a time domain waveform chart of each electric signal time subsequence;
taking the data mapping result of the information entropy and standard deviation product of all elements in the trough change characteristic value sequence of the time domain waveform diagram of each electric signal time sub-sequence as a molecule;
taking the sum of the absolute values of the difference values of all adjacent elements in the trough change characteristic value sequence of the time domain waveform diagram of each electric signal time subsequence and 1 as a denominator;
Taking the difference value of the ratio of 1 to the numerator and the denominator as the ripple noise interference trough index of the corresponding time period of each electrical signal time subsequence.
Preferably, the method for determining the long-term disturbance coefficient of the measured disturbance in the load power supply measurement process based on the ripple noise interference joint index of all time periods and the consistency characteristic of the long-term autocorrelation among the electrical signal time subsequences comprises the following steps:
determining the length Cheng Wending coefficient of the electric signal of each time period based on the correlation strength between the trending sequences of the electric signal time sub-sequences in all the time periods and the stability degree of the long-range autocorrelation;
Determining noise influence complexity of the electrical signal time sequence based on clustering results of the ripple noise interference joint indexes of all time periods;
The sum of the electric signal length Cheng Wending coefficients of all time periods is taken as a denominator, and the ratio of the noise influence complexity of the electric signal time sequence to the denominator is taken as a long-term disturbance coefficient of the measurement interference of the electric signal time sequence.
Preferably, the method for determining the length Cheng Wending coefficient of the electrical signal in each time period based on the correlation strength between the detrending sequences of the electrical signal time sub-sequences in all time periods and the stability degree of the long-range autocorrelation is as follows:
Taking the electrical signal time sub-sequence in each time period as input, and acquiring a trending sequence corresponding to the electrical signal time sub-sequence in each time period by adopting a DFA algorithm;
Taking a sequence formed by absolute values of differences between all the element values in the same sequence in the corresponding trend removing sequence of each electrical signal time subsequence according to time sequence as a measurement long-range stable sequence of each electrical signal time subsequence;
Taking a similarity measurement result between the electrical signal time sub-sequence in each time period and the measured long-range stable sequence of the electrical signal time sub-sequence in any one of the rest time periods as a molecule;
taking the difference value between the maximum value in the measured long-range stable sequence of the electrical signal time subsequence in each time period and the average value of all elements as a first difference value, and taking the sum of the product of the first difference value and the average value and 0.1 as a denominator;
The average value of the accumulated results of the ratio of the numerator to the denominator over all the time periods is taken as the electric signal length Cheng Wending coefficient of each time period.
Preferably, the method for determining the noise influence complexity of the electrical signal time sequence based on the clustering result of the ripple noise interference joint index of all time periods comprises the following steps:
Taking the ripple noise interference joint indexes of all time periods as input, and adopting a data clustering algorithm to obtain a clustering result of the ripple noise interference joint indexes;
and calculating variation coefficients of the ripple noise interference joint indexes in all time periods, and taking the product of the accumulation results of Euclidean distances between the central points of any two clusters in the clustering results on all clusters and the variation coefficients as the noise influence complexity of the electrical signal time sequence.
Preferably, the method for determining the electrical signal ripple abnormality index of the electrical signal time sequence based on the ripple noise interference combination index of each time period and the measured interference long-term disturbance coefficient in the load point power supply measuring process comprises the following steps:
taking the data mapping result of the disturbance coefficient of the electric signal time sequence with the long-term disturbance of the measurement as a molecule;
taking the sum of the average value and 1 of the ripple noise interference joint indexes of all time periods as a denominator;
and taking the difference value of the ratio of 1 to the numerator and the denominator as an electrical signal ripple abnormality index of the electrical signal time sequence.
Preferably, the method for determining the step factor of each iteration in the electric signal filtering process based on the electric signal ripple abnormal index of the electric signal time sequence comprises the following steps:
And taking the product of the reciprocal of the maximum eigenvalue of the autocorrelation matrix at each iteration in the LMS filtering algorithm and the electrical signal ripple abnormal index of the electrical signal time sequence at each iteration as a step factor of each iteration in the filtering process.
Preferably, the method for obtaining the performance measurement result of the load point power supply by adopting the LMS filtering algorithm based on the step factor comprises the following steps:
taking the electrical signal time sequence as input, and obtaining a pure electrical signal sequence corresponding to the electrical signal time sequence based on the step factor by adopting an LMS filtering algorithm;
and taking a pure electric signal sequence corresponding to the electric signal time sequence as input, converting the pure electric signal sequence into a current time sequence by adopting a current converter, taking the current time sequence as input, and obtaining a stability measurement result of the load point power supply by adopting an ADF (automatic frequency filter) inspection algorithm.
The beneficial effects of the application are as follows: the wave crest change characteristic value sequence and the wave trough change characteristic value sequence are obtained according to the electric signal time sequence, the ripple noise interference joint index is constructed according to the wave crest change characteristic value sequence and the wave trough change characteristic value sequence, and the interference degree of the ripple noise in the power supply loop on the electric signal time subsequence in different time periods can be reflected more carefully; secondly, the long-range autocorrelation consistency characteristics of the electrical signal time sequences in different time periods are combined to construct a disturbance long-time disturbance coefficient for measuring disturbance, the disturbance long-time disturbance coefficient for measuring disturbance considers the trend consistency characteristics among electrical signal time subsequences in different time periods, and the influence of different measurement data on disturbance degree evaluation in the performance measurement process can be reduced; secondly, based on the long-term disturbance coefficient of measurement interference and the ripple noise interference joint index, the noise influence complexity is obtained, the abnormal fluctuation index of the electric signal is obtained according to the noise influence complexity, and the step length factor is adaptively adjusted according to the abnormal fluctuation index of the electric signal.
Drawings
In order to more clearly illustrate the embodiments of the application or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the application, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
FIG. 1 is a flow chart of a method for measuring performance of a point-of-load power supply according to an embodiment of the application;
Fig. 2 is a flowchart of an implementation of a method for measuring performance of a point-of-load power supply according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Referring to fig. 1, a flowchart of a method for measuring performance of a point-of-load power supply according to an embodiment of the application is shown, the method includes the following steps:
and S001, collecting an electric signal in a power supply loop of the load point power supply.
The application aims to reduce the interference of ripple noise on an electric signal by digitally filtering a current signal in a power supply loop of a load point power supply, obtain current data in the power supply loop of the load point power supply through signal conversion, and obtain a performance measurement result of the load point power supply by measuring the stability of the current in the power supply loop of the load point power supply.
Specifically, the electric signal sensor is used for collecting the electric signal in the power supply loop of the load point power supply, the bit width of the electric signal is 16 bits, the sampling frequency is 8KHz, the total sampling time is 1h, and an operator can set proper sampling bit width and sampling frequency according to the actual condition of the load point power supply. And regarding the acquired electric signals, taking a sequence formed by all the acquired electric signals according to the ascending order of time as an electric signal time sequence.
Thus, the time sequence of the electric signal in the measuring process of the load point power supply is obtained and is used as the object of the subsequent signal filtering.
Step S002, based on the evaluation result of the interference degree of the extreme point in the time domain waveform diagram of the electrical signal time sub-sequence in each time period, determining the ripple noise interference joint index in each time period.
The ripple noise in the power supply loop produces larger interference on the electric signal, so that the accuracy of the electric signal is higher, and the accuracy of the current obtained by converting the electric signal by using a signal conversion technology is lower, so that the accuracy of the measurement result of the power supply performance of the load point in the power supply loop is lower. When the LMS filtering algorithm is used for filtering and denoising the electric signal, the contradiction between the convergence speed and steady state offset can influence the precision of time delay estimation, thereby influencing the denoising effect on ripple noise in the electric signal and further leading to lower accuracy of the measurement of the power performance of the load point.
Specifically, since the influence degree of ripple noise on the electric signal in different time periods in the power supply loop of the load point power supply is different, in order to more finely analyze the influence of the ripple noise on the electric signal, a uniform segmentation mode is adopted in a time dimension, the electric signal time sequence is divided into K time periods, each time period corresponds to one electric signal time sub-sequence, and the empirical value of K is 30. Taking the electrical signal time sub-sequence in the ith time period as an example, the electrical signal time sub-sequence in the ith time period is taken as input, and a time domain waveform diagram of the electrical signal time sub-sequence in the ith time period is drawn by MATLAB software, wherein the MATLAB software drawing is a known technology, and the specific process is not repeated.
Further, in the time domain waveform of the electric signal in each period, fluctuation variation of the electric signal is generally uniform, but variation of the electric signal occurs due to influence of noise components. In order to identify the difference characteristic of the variation of the electric signal, for a time domain waveform diagram of the electric signal time subsequence in the ith time period, a set formed by L wave crest points with the nearest Euclidean distance between each wave crest point in the time domain waveform diagram is used as a neighboring wave crest point set of each wave crest point; and taking a set formed by L trough points closest to the Euclidean distance between every two trough points as a neighboring trough point set of each trough point, wherein the empirical value of L is 2.
And secondly, regarding any one peak point in the time domain waveform diagram of the electrical signal time sub-sequence in the ith time period, taking the mean value of Euclidean distances between each peak point and all peak points in the adjacent peak point set of each peak point as the variation characteristic value of each peak point, respectively calculating the variation characteristic values of all the peak points, and taking a sequence formed by the variation characteristic values of all the peak points according to the ascending order of time as the peak variation characteristic value sequence of the time domain waveform diagram of the electrical signal time sub-sequence in the ith time period. Similarly, for any one wave trough point in the time domain waveform diagram of the electrical signal time sub-sequence in the ith time period, taking the mean value of Euclidean distances between each wave trough point and all wave trough points in the adjacent wave trough point set of each wave trough point as the change characteristic value of each wave trough point, respectively calculating the change characteristic values of all wave trough points, and taking a sequence formed by the change characteristic values of all wave trough points according to the ascending order of time as the wave trough change characteristic value sequence of the time domain waveform diagram of the electrical signal time sub-sequence in the ith time period. The peak change characteristic value sequence and the trough change characteristic value sequence reflect the severity of interference of extreme points in the electrical signal time subsequence to a certain extent.
Based on the above analysis, a ripple noise interference joint index is constructed here to characterize the degree to which the time sub-sequence of electrical signals within each time period is interfered by the ripple noise in the power supply loop. Calculating the ripple noise interference joint index of the ith time period:
，
，
，
in the method, in the process of the invention, Is the ripple noise interference crest index of the i-th period,、The information entropy and the average value of all elements in the peak change characteristic value sequence corresponding to the electric signal time subsequence in the ith time period respectively,Is an exponential function based on a natural constant, N is the number of elements in a peak change characteristic value sequence corresponding to the electrical signal time subsequence in the ith time period,、The j-1 elements in the peak change characteristic value sequence corresponding to the electrical signal time subsequence in the ith time period are respectively;
is the ripple noise floor index of the i-th time period, 、The information entropy and the average value of all elements in the trough change characteristic value sequence corresponding to the electric signal time subsequence in the ith time period are respectively, M is the number of elements in the trough change characteristic value sequence corresponding to the electric signal time subsequence in the ith time period,、The s-1 element in the trough change characteristic value sequence corresponding to the electric signal time subsequence in the ith time period;
Is the ripple noise interference combination index for the i-th time period, 、The respective weights are the fusion weights and,、The empirical values of 0.5 and 0.5 are respectively adopted.
Wherein the greater the degree of influence of ripple noise in the power supply loop on the time sub-sequence of the electric signal in the ith time period, the greater the Euclidean distance difference between adjacent peak points, the greater the difference between the variation characteristic values of the peak points,The larger the value of the (i) is, the higher the degree of dispersion of the element distribution in the peak change characteristic value sequence corresponding to the electrical signal time subsequence in the ith time period is, the larger the information entropy and standard deviation of the element in the peak change characteristic value sequence are,、The larger the value of (c) is,The smaller the value of (2); similarly, the greater the influence degree of ripple noise in the power supply loop on the electrical signal time subsequence in the ith time period, the greater the Euclidean distance difference between adjacent wave trough points, the greater the difference between the change characteristic values of the wave trough points, the higher the discrete degree of element distribution in the wave trough change characteristic value sequence corresponding to the electrical signal time subsequence in the ith time period, the higher the information entropy and standard deviation of elements in the wave trough change characteristic value sequence,The larger the value of (c) is,、The larger the value of (c) is,The smaller the value of (2); i.e.The greater the value of (2), the more severely the time sub-sequence of electrical signals within the ith time period is disturbed by ripple noise in the supply loop.
Thus, the ripple noise interference joint index of each time period is obtained and is used for subsequently evaluating the interference degree of the electric signal in the measuring process.
Step S003, determining a long-term disturbance coefficient of measurement interference in the load power supply measurement process based on the ripple noise interference combination index and the consistency characteristic of long-range autocorrelation among the electrical signal time subsequences; and determining the electrical signal ripple abnormality index based on the ripple noise interference joint index and the measured interference long-term disturbance coefficient.
In order to obtain the influence of the ripple noise on the whole electric signal time sequence, the complexity of the influence of the ripple noise on the electric signal time sequence is further analyzed. Because the influence of the ripple noise on the electric signals in different time periods is different, when the influence degree of the ripple noise on the electric signal time sub-sequences in different time periods is large, the influence complexity of the ripple noise on the electric signal time sub-sequences in all time periods is high.
Specifically, according to the steps, the ripple noise interference combination indexes of each time period are calculated respectively, the ripple noise interference combination indexes of all time periods are taken as input, a density peak clustering DPC (DENSITY PEAKS clustering) algorithm is adopted to obtain a clustering result of the ripple noise interference combination indexes, the clustering result comprises different clustering clusters, the DPC clustering algorithm is a known technology, and the specific process is not repeated.
In another embodiment, for the combined indexes of the ripple noise and the interference in all the time periods, the combined indexes of the ripple noise and the interference in all the time periods may be used as input, and a k-means clustering algorithm is used to obtain a clustering result of the combined indexes of the ripple noise and the interference, where the k-means clustering algorithm is a known technology, and a specific process is not repeated.
Further, if the performance of the load point power supply is better and the measurement process is not affected by noise interference, the time intervals between the sampling points corresponding to different amplitudes should be equal in a plurality of electric signal periods of the time domain waveform diagram, that is, the long-term autocorrelation in the electric signal time sequence has higher stability, and the long-term autocorrelation in each time period is almost consistent. Thus, for any one time period, taking the ith time period as an example, the electrical signal time in the ith time period is sub-sequencedAs an input to the detrend DFA (Detrended fluctuation analysis) algorithm, the DFA algorithm is used to obtainIs a trending sequence of (2)The DFA algorithm is a well-known technique, and the specific process is not described in detail. Second, calculate、Any same position between the two elements corresponds to the absolute value of the difference between the two elements、The absolute value of the difference between two elements corresponding to all the same positions is taken as the sequence formed by time sequenceIs a measurement of long-range stable sequences.
Based on the analysis, a disturbance coefficient for measuring the long-term disturbance of the disturbance is constructed and used for representing the complexity of the electric signal time sequence affected by the disturbance in different time periods. The specific calculation formula for measuring the disturbance factor of the long-term disturbance is as follows:
，
，
，
in the method, in the process of the invention, Is the electric signal length Cheng Wending coefficient of the i-th time period, K is the number of time periods in which the electric signal time series is divided,、The measurement long-range stable sequences of the ith and jth time periods respectively,Is a sequence of、The pearson correlation coefficient between them,、Respectively are sequences ofThe average value of all elements in (c), the maximum value of all elements,Is a parameter adjusting factor for preventing denominator from being 0,The magnitude of (2) is 0.1, the pearson correlation coefficient is a known technology, and the specific process is not repeated;
c is the noise-affecting complexity of the electrical signal time series, For the variation coefficient of the ripple noise interference combination index of all time periods, m is the number of clusters in the clustering result of the ripple noise interference combination index of all time periods,、Respectively the first of the clustering resultsFirst, secondThe number of clusters of the clusters is,Is a function of the euclidean distance,AndRespectively represent the first in the clustering resultFirst, secondThe location of the center point within the cluster of clusters,Represent the firstFirst, secondThe Euclidean distance between the center point positions in the cluster clusters;
V is the disturbance coefficient of the measurement of the electrical signal time series during the long-term disturbance.
The more complex the influence of the ripple noise on the electrical signal time sequence, the more inconsistent the influence degree of the ripple noise on the electrical signal time sub-sequences in different time periods, the more unstable the autocorrelation degree of the electrical signal time sequence in different time periods, the worse the long-range autocorrelation, and the electrical signal time sub-sequences in the ith time periodAnd detrending sequences thereofThe larger the difference is, the less concentrated the element distribution in the corresponding measured long-range stable sequence is, the first differenceThe larger the values of the sequences are, the more inconsistent the interference influence is received by the time sub-sequences of the electric signals in different time periods, the larger the trend difference of the time sub-sequences of the electric signals in different time periods is, the larger the difference between the corresponding detrending sequences is, the smaller the similarity is between the measured long-range stable sequences in different time periods,The smaller the value of (c) is,The smaller the value of (2); the more inconsistent the degree of influence of the ripple noise on the time sub-sequences of the electrical signal in different time periods, the more inconsistent the magnitude of the ripple noise interference joint index in different time periods,The larger the value of (2), the larger the inter-class difference between different clusters in the clustering result of all the ripple noise interference joint indexes,The larger the value of C; i.e. the larger the value of V, the greater the complexity of the influence of the ripple noise on the time series of the electrical signal, the greater the complexity of the noise influence.
Further, the fluctuation degree of the electric signal in the measuring process is comprehensively estimated based on the ripple noise interference joint indexes of all time periods and the long-term disturbance coefficients of the measuring interference of the electric signal time sequence.
，
Wherein T is an electrical signal ripple anomaly index of the electrical signal time sequence,Is an exponential function based on a natural constant, V is a long-term disturbance coefficient of the measurement interference of the electric signal time sequence,Is the average of the ripple noise interference combination index for all time periods.
Wherein the average value of the ripple noise interference combination indexThe larger the influence of the ripple noise on the electric signal time sequence is, the more noise components are contained in the electric signal, the larger the possibility of abnormal fluctuation phenomenon of the electric signal is, the larger the value of the disturbance coefficient V is measured at the same time, the larger the influence complexity of the ripple noise on the electric signal time sequence is, the larger the possibility of abnormality of the electric signal is, and the larger the value of the electric signal ripple abnormality index T is.
Thus, an electrical signal ripple abnormality index of the electrical signal time series is obtained.
Step S004, determining a step factor of each iteration in the filtering process based on the electrical signal ripple abnormality index of the electrical signal time sequence; and obtaining a performance measurement result of the load point power supply based on the step factor by adopting an LMS filtering algorithm.
The LMS adaptive filtering algorithm is an iterative optimization filtering algorithm, each iteration of the LMS adaptive filtering algorithm can obtain an output electric signal time sequence, and through the steps, each iteration of the LMS adaptive filtering algorithm can obtain an electric signal ripple abnormal index of the electric signal time sequence. Secondly, a step factor at each iteration is adaptively determined based on the electrical signal ripple anomaly index.
Specifically, step size factor in conventional LMS adaptive filtering algorithmThe range of the values is as followsWhereinAutocorrelation matrix for the t-th iterationIs the maximum eigenvalue of (c). The application calculates the step length factor of each iteration in the LMS self-adaptive filtering algorithm based on the electrical signal ripple abnormality index of the electrical signal time sequence of each iteration. The step factor at the t-th iteration is calculated as follows:
，
in the method, in the process of the invention, A step size factor representing the t-th iteration of the LMS adaptive filtering algorithm,Autocorrelation matrix representing the t-th iteration in LMS adaptive filtering algorithmIs used for the maximum characteristic value of the (c),And the electrical signal ripple abnormality index represents the electrical signal time sequence at the t-th iteration in the LMS adaptive filtering algorithm. Wherein,The calculation of (2) is a well-known technique in the LMS algorithm, and the specific process is not repeated.
Wherein, the ripple noise component in the time sequence of the electric signal is more in the t iteration and the state of steady state convergence is not reached, a larger step factor should be selected at the moment to quickly reach the state of steady state convergence,The larger the value of (2), the step factorThe greater the value of (2); the more the ripple abnormality index of the electric signal approaches 0, which means that the ripple noise component in the time sequence of the electric signal is less at this time, and approaches to a state of steady state convergence at this time,The smaller the value of (2), the smaller step factor should be selected at this time to ensure the stability of the steady state convergence error and avoid the divergence phenomenon of the steady state.
According to the steps, step factors at each iteration are calculated respectively. Secondly, taking the time sequence of the electric signal as input, taking the step factor at each iteration as the step feedback factor in the algorithm, and obtaining a pure electric signal sequence corresponding to the time sequence of the electric signal by using an LMS filtering algorithm, wherein the LMS filtering algorithm is a known technology, and the specific process is not repeated. Further, taking a pure electric signal sequence corresponding to the electric signal time sequence as input, converting the pure electric signal sequence into a current time sequence by adopting a current converter, taking the current time sequence as input, and obtaining a stability measurement result of the load point power supply by adopting an ADF (automatic digital-fuse) test algorithm, wherein when the stability test result is stable, the performance measurement result of the load point power supply is excellent; when the stability test result is non-stable, the performance measurement result of the load point power supply is weak, wherein the use of the current converter and the ADF test algorithm are both known techniques, and the specific process is not repeated. The resulting stationarity measurements and their corresponding performance measurements are then recorded in a measurement database, the overall measurement flow being shown in fig. 2.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. The foregoing description of the preferred embodiments of the present application is not intended to be limiting, but rather, any modifications, equivalents, improvements, etc. that fall within the principles of the present application are intended to be included within the scope of the present application.
Claims (10)
1. A method for measuring performance of a point-of-load power supply, the method comprising the steps of:
collecting an electric signal in a load point power supply loop;
Determining a ripple noise interference joint index of each time period based on an evaluation result of the interference degree of the extreme point in the time domain waveform diagram of the electrical signal time sub-sequence in each time period;
Determining a long-term disturbance coefficient of the measured interference in the load power supply measuring process based on the ripple noise interference joint index of all time periods and the consistency characteristic of the long-range autocorrelation among the electrical signal time subsequences;
determining an electrical signal ripple abnormality index of the electrical signal time sequence based on the ripple noise interference joint index of each time period and the measured interference long-term turbulence coefficient in the load point power supply measuring process;
Determining a step factor of each iteration in the electric signal filtering process based on the electric signal ripple abnormal index of the electric signal time sequence; and obtaining a performance measurement result of the load point power supply based on the step factor by adopting an LMS filtering algorithm.
2. The method for measuring the performance of a load point power supply according to claim 1, wherein the method for determining the ripple noise interference joint index of each time period based on the evaluation result of the interference degree of the extreme point in the time domain waveform of the time sub-sequence of the electric signal in each time period is as follows:
Taking a sequence formed by all the collected electric signals according to the ascending order of time as an electric signal time sequence, and dividing the electric signal time sequence into a preset number of electric signal time subsequences in a time dimension in a uniform dividing mode;
Respectively taking each electrical signal time sub-sequence as input, and acquiring a time domain waveform diagram of each electrical signal time sub-sequence by utilizing MATLAB software;
determining a ripple noise interference crest index of each time period based on the stability degree of the euclidean distance between the crest points in the time domain waveform diagram of each electrical signal time sub-sequence;
determining a ripple noise interference trough index of each time period based on the stability degree of the euclidean distance between trough points in the time domain waveform diagram of each electrical signal time sub-sequence;
The ripple noise interference combined index of each time period consists of a ripple noise interference crest index and a ripple noise interference trough index, wherein the ripple noise interference combined index is respectively in positive correlation with the ripple noise interference crest index and the ripple noise interference trough index.
3. The method for measuring the performance of a load point power supply according to claim 2, wherein the method for determining the ripple noise interference crest index of each time period based on the stability of the euclidean distance between the crest points in the time domain waveform of each electrical signal time sub-sequence is as follows:
Taking a set formed by a preset number of peak points with the minimum Euclidean distance with each peak point in a time domain oscillogram of each electric signal time sub-sequence as an adjacent peak point set of each peak point;
Taking the mean value of Euclidean distances between each peak point and all the peak points in the adjacent peak point set of each peak point as the variation characteristic value of each peak point, and taking a sequence formed by the variation characteristic values of all the peak points according to the sequence of time ascending as the peak variation characteristic value sequence of the time domain waveform chart of each electric signal time subsequence;
taking the data mapping result of the information entropy and standard deviation product of all elements in the wave crest variation characteristic value sequence of the time domain waveform chart of each electric signal time sub-sequence as a molecule;
taking the sum of the absolute values of the difference values of all adjacent elements in the wave crest variation characteristic value sequence of the time domain waveform diagram of each electric signal time subsequence and 1 as a denominator;
Taking the difference value of the ratio of 1 to the numerator and the denominator as the ripple noise interference crest index of the corresponding time period of each electrical signal time subsequence.
4. The method for measuring the performance of a load point power supply according to claim 2, wherein the method for determining the ripple noise interference trough index of each time period based on the stability of the euclidean distance between the trough points in the time domain waveform of each electrical signal time sub-sequence is as follows:
Taking a set formed by a preset number of wave crest points with the minimum Euclidean distance with each wave trough point in a time domain oscillogram of each electric signal time subsequence as an adjacent wave trough point set of each wave trough point;
Taking the average value of Euclidean distances between each wave valley point and all wave valley points in a nearby wave valley point set of each wave valley point as a change characteristic value of each wave valley point, and taking a sequence formed by the change characteristic values of all wave valley points according to the sequence of time ascending as a wave valley change characteristic value sequence of a time domain waveform chart of each electric signal time subsequence;
taking the data mapping result of the information entropy and standard deviation product of all elements in the trough change characteristic value sequence of the time domain waveform diagram of each electric signal time sub-sequence as a molecule;
taking the sum of the absolute values of the difference values of all adjacent elements in the trough change characteristic value sequence of the time domain waveform diagram of each electric signal time subsequence and 1 as a denominator;
Taking the difference value of the ratio of 1 to the numerator and the denominator as the ripple noise interference trough index of the corresponding time period of each electrical signal time subsequence.
5. The method for measuring the performance of a load point power supply according to claim 1, wherein the method for determining the long-term disturbance factor of the measured disturbance in the load power supply measurement process based on the ripple noise interference combination index of all time periods and the consistency characteristic of the long-term autocorrelation among the electrical signal time sub-sequences is as follows:
determining the length Cheng Wending coefficient of the electric signal of each time period based on the correlation strength between the trending sequences of the electric signal time sub-sequences in all the time periods and the stability degree of the long-range autocorrelation;
Determining noise influence complexity of the electrical signal time sequence based on clustering results of the ripple noise interference joint indexes of all time periods;
The sum of the electric signal length Cheng Wending coefficients of all time periods is taken as a denominator, and the ratio of the noise influence complexity of the electric signal time sequence to the denominator is taken as a long-term disturbance coefficient of the measurement interference of the electric signal time sequence.
6. The method for measuring the performance of a point-of-load power supply according to claim 5, wherein the method for determining the length Cheng Wending coefficient of the electrical signal in each time zone based on the correlation strength between the detrending sequences of the electrical signal time sub-sequences and the stability of the long-range autocorrelation in all time zones is as follows:
Taking the electrical signal time sub-sequence in each time period as input, and acquiring a trending sequence corresponding to the electrical signal time sub-sequence in each time period by adopting a DFA algorithm;
Taking a sequence formed by absolute values of differences between all the element values in the same sequence in the corresponding trend removing sequence of each electrical signal time subsequence according to time sequence as a measurement long-range stable sequence of each electrical signal time subsequence;
Taking a similarity measurement result between the electrical signal time sub-sequence in each time period and the measured long-range stable sequence of the electrical signal time sub-sequence in any one of the rest time periods as a molecule;
taking the difference value between the maximum value in the measured long-range stable sequence of the electrical signal time subsequence in each time period and the average value of all elements as a first difference value, and taking the sum of the product of the first difference value and the average value and 0.1 as a denominator;
The average value of the accumulated results of the ratio of the numerator to the denominator over all the time periods is taken as the electric signal length Cheng Wending coefficient of each time period.
7. The method for measuring the performance of a load point power supply according to claim 5, wherein the method for determining the noise influence complexity of the electrical signal time sequence based on the clustering result of the ripple noise interference combination index of all time periods is as follows:
Taking the ripple noise interference joint indexes of all time periods as input, and adopting a data clustering algorithm to obtain a clustering result of the ripple noise interference joint indexes;
and calculating variation coefficients of the ripple noise interference joint indexes in all time periods, and taking the product of the accumulation results of Euclidean distances between the central points of any two clusters in the clustering results on all clusters and the variation coefficients as the noise influence complexity of the electrical signal time sequence.
8. The method for measuring the performance of a load point power supply according to claim 1, wherein the method for determining the electrical signal ripple anomaly index of the electrical signal time sequence based on the ripple noise interference joint index of each time period and the measured interference long-term turbulence coefficient in the load point power supply measurement process comprises the following steps:
taking the data mapping result of the disturbance coefficient of the electric signal time sequence with the long-term disturbance of the measurement as a molecule;
taking the sum of the average value and 1 of the ripple noise interference joint indexes of all time periods as a denominator;
and taking the difference value of the ratio of 1 to the numerator and the denominator as an electrical signal ripple abnormality index of the electrical signal time sequence.
9. The method for measuring the performance of a point-of-load power supply according to claim 1, wherein the method for determining the step factor of each iteration in the electric signal filtering process based on the electric signal ripple abnormality index of the electric signal time sequence is as follows:
And taking the product of the reciprocal of the maximum eigenvalue of the autocorrelation matrix at each iteration in the LMS filtering algorithm and the electrical signal ripple abnormal index of the electrical signal time sequence at each iteration as a step factor of each iteration in the filtering process.
10. The method for measuring the performance of the load point power supply according to claim 1, wherein the method for obtaining the performance measurement result of the load point power supply based on the step factor by using the LMS filtering algorithm is as follows:
taking the electrical signal time sequence as input, and obtaining a pure electrical signal sequence corresponding to the electrical signal time sequence based on the step factor by adopting an LMS filtering algorithm;
and taking a pure electric signal sequence corresponding to the electric signal time sequence as input, converting the pure electric signal sequence into a current time sequence by adopting a current converter, taking the current time sequence as input, and obtaining a stability measurement result of the load point power supply by adopting an ADF (automatic frequency filter) inspection algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410815436.XA CN118376948B (en) | 2024-06-24 | 2024-06-24 | Performance measurement method of load point power supply |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410815436.XA CN118376948B (en) | 2024-06-24 | 2024-06-24 | Performance measurement method of load point power supply |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118376948A CN118376948A (en) | 2024-07-23 |
| CN118376948B true CN118376948B (en) | 2024-08-23 |
Family
ID=91904269
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410815436.XA Active CN118376948B (en) | 2024-06-24 | 2024-06-24 | Performance measurement method of load point power supply |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118376948B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118708924B (en) * | 2024-08-27 | 2024-12-13 | 卡尔德线缆(东莞)有限公司 | A robot cable electrical performance testing method, device and system |
| CN119291382B (en) * | 2024-10-18 | 2025-05-30 | 广东思科通用电力科技有限公司 | Bus duct fault monitoring method and system |
| CN119298940B (en) * | 2024-12-11 | 2025-04-01 | 北京星光裕华光电科技有限公司 | Power line carrier signal transmission method and device applied to direct current lighting system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113424422A (en) * | 2019-03-29 | 2021-09-21 | 三垦电气株式会社 | Switching power supply device |
| CN116454079A (en) * | 2023-03-27 | 2023-07-18 | 天航长鹰(江苏)科技有限公司 | High-performance load point power supply |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008165583A (en) * | 2006-12-28 | 2008-07-17 | Fujitsu Ltd | Circuit system, circuit unit, power supply unit, and power supply method |
| US10917009B2 (en) * | 2016-08-09 | 2021-02-09 | California Institute Of Technology | Digital multiphase hysteretic point-of-load DC/DC converter |
-
2024
- 2024-06-24 CN CN202410815436.XA patent/CN118376948B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113424422A (en) * | 2019-03-29 | 2021-09-21 | 三垦电气株式会社 | Switching power supply device |
| CN116454079A (en) * | 2023-03-27 | 2023-07-18 | 天航长鹰(江苏)科技有限公司 | High-performance load point power supply |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118376948A (en) | 2024-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN118376948B (en) | Performance measurement method of load point power supply | |
| CN116609720B (en) | Data-driven-based intelligent error compensation method and system for desk-top multimeter | |
| CN118228006B (en) | A chip detection method and system based on FPGA technology | |
| CN118294786B (en) | Chip detection method and system based on artificial intelligence | |
| CN117783658B (en) | A digital current measurement method for electric power equipment | |
| CN118070195A (en) | Mining alternating current frequency converter abnormal data state monitoring system | |
| CN114062993A (en) | CVT error state prediction method based on time convolution network | |
| CN110895626B (en) | Performance degradation model precision verification method based on leave-one-out cross verification | |
| CN118740114B (en) | Current digital filtering method of power equipment | |
| CN117129904A (en) | Industrial power supply rapid switching monitoring method based on data analysis | |
| CN118376953B (en) | Load point power interruption test method and system | |
| CN118152829B (en) | Health state assessment method and system for oil type iron core reactor | |
| CN117171503B (en) | Intelligent prediction method for battery charging time | |
| Chang et al. | Micro-short circuit fault diagnosis of lithium-ion battery based on voltage curve similarity ranking volatility | |
| CN118033409A (en) | GCB arc extinguishing chamber switch resistance testing method | |
| KR20210129453A (en) | Apparatus and method of detecting bad data in power system measurement data | |
| CN113343877B (en) | Electrical appliance fingerprint feature extraction method based on different Prony methods | |
| CN118708924B (en) | A robot cable electrical performance testing method, device and system | |
| CN119917982B (en) | Operation fault detection method and system for HIAF synchronous timing system | |
| CN119125688B (en) | Method and device for accurately measuring parameters of passive filter and electronic equipment | |
| CN116679165B (en) | Frequency domain reflection cable defect positioning method based on synchronous extrusion generalized S transformation | |
| CN118707286B (en) | Diode electrical characteristic testing method | |
| CN119003302B (en) | Steady state detection method and device for SSD device and SSD device | |
| CN120104974A (en) | Isolation switch health state management method and system based on AI | |
| CN120177860A (en) | Overvoltage early warning method and system in state of new energy being connected into power distribution network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |